太赫兹科学与电子信息学报2026，Vol.24Issue(6)：721-727,7.DOI:10.11805/TKYDA2025099
基于1D-CNN的已知未知混合雷达信号识别
Identification of known unknown hybrid radar signals based on 1D-CNN
付懿洋 1潘继飞 2刘志伟1
作者信息
- 1. 91715部队,广东 广州 510000
- 2. 国防科技大学 电子对抗学院,安徽 合肥 230037
- 折叠
摘要
Abstract
Radar signal recognition is one of the critical tasks in modern radar electronic warfare systems.Traditional radar signal recognition methods primarily rely on experienced operators and sophisticated signal processing techniques.With the continuous advancement of radar technology,the quantity and variety of radar signals in the electromagnetic environment are increasing and overlapping with each other.When identifying mixed known and unknown radar signals,conventional recognition methods often suffer from low recognition accuracy and poor generalization capability.This paper proposes a method based on one-Dimensional Convolutional Neural Network(1D-CNN)for recognizing mixed known and unknown radar signals.Through recognition experiments on three simulated Pulse Description Word(PDW)datasets of mixed known and unknown radar signals with varying complexity levels,the recognition accuracy reaches no less than 89.5%,which verifies the feasibility of the proposed method.关键词
雷达信号识别/已知未知混合雷达信号/一维卷积神经网络/脉冲描述字(PDW)Key words
radar signal identification/known unknown hybrid radar signal/one-Dimensional Convolutional Neural Network(1D-CNN)/Pulse Description Word(PDW)分类
信息技术与安全科学引用本文复制引用
付懿洋,潘继飞,刘志伟..基于1D-CNN的已知未知混合雷达信号识别[J].太赫兹科学与电子信息学报,2026,24(6):721-727,7.
